In documents released Monday ahead of an FDA advisory panel meeting scheduled for September 14, agency staff raised concerns over a post-marketing study assessing the neuropsychiatric effects of Pfizer's smoking cessation product Chantix (varenicline). Reviewers noted that inconsistencies in data collection and characterisation of the severity of adverse events may have biased the study results in favour of Chantix.
In 2009, based on a review of adverse-event reports, the FDA required both Chantix and GlaxoSmithKline's smoking cessation treatment Zyban (bupropion) to carry boxed warnings  about the risk of serious neuropsychiatric events. The regulator also requested that the companies conduct a post-marketing study "to further characterise the risk of neuropsychiatric adverse events and to evaluate whether a prior history of psychiatric illness was a risk modifier." Pfizer has been trying to get the boxed warning removed from the Chantix label, although an FDA advisory panel in 2014 recommended that it be kept, while agency staff said at the time that no decision should be made regarding the removal of the warning until the post-marketing requirement is completed.
FDA reviewers said results from the now-completed trial of about 8000 participants showed that among patients who had already been diagnosed with a psychiatric disorder, the observed rate of neuropsychiatric events was 6.5 percent and 6.7 percent for those treated with Chantix and Zyban, respectively, compared with 4.9 percent for the placebo group. Meanwhile, among those without a psychiatric history diagnosis at baseline, the rate was 1.3 percent for the Chantix arm, versus 2.2 percent for the Zyban group and 2.4 percent for placebo.
In the report, agency staff said Pfizer's "ascertainment of the primary neuropsychiatric adverse event outcomes" might have impacted the interpretation of the trial's results. Moreover, the FDA suggested Pfizer did not properly use interview techniques, noting there was "substantial variability among investigators" in how adverse events were recorded, as well as "inconsistency" in the handling of suicidality cases. FDA staff concluded that these "factors likely served to lower the overall number of primary outcome events" and could have led to biased results.
In response, James Rusnak, chief development officer of Pfizer's cardiovascular metabolic unit, noted that there were bound to be variations in investigators' judgment within the boundaries of the trial protocol. "Ultimately, what was reported as an adverse event, was determined based upon the investigator's local determination of the subject presenting in front of them," Rusnak explained, adding that any subjectivity in investigator's judgment would be spread across the four treatment groups and not be limited to Chantix.
Reviewers also questioned whether researchers who had financial ties with Pfizer may have underreported side effects in the study. Specifically, the agency analysis uncovered that investigators at 32 trial sites had received payments from Pfizer of at least $25 000 for speaking on behalf of Chantix or other company products. "Overall, sites without financial involvement observed a higher percentage of neuropsychiatric events than sites with financial involvement" in both of the study's cohorts, the agency said, but cautioned that the findings could be due to chance given the small sample size.
Meanwhile, regulators in Europe, where Pfizer markets the therapy under the name Champix, removed  a warning on the product's label earlier this year regarding the potential for increased risk of neuropsychiatric adverse events.